The adjective ‘uncanny’ describes something that is not only enigmatic but also somehow unsettling because of its familiarity. In 1970, Japanese roboticist Masahiro Mori utilised the term in his essay ‘Bukimi no Tani’ – that is, ‘The Uncanny Valley’ – to caution robot builders against making their creations resemble humans too closely, to avoid the possible terrors of likeness. Such warnings are taken in vain by composer Daniel M. Karlsson in his latest album, Mapping the Valleys of the Uncanny.

The whole album has been composed with algorithms that produce structures which are populated with acoustic sounds that have been heavily processed. To that end, the disc is supposed to be ‘uncanny’ and, presumably, resemble contemporary classical music, which it does very effectively, and there are undoubtedly some very nice, perhaps even beautiful, sounds here. The album opener, ‘Intersections of traversals’, presents a 40-minute-long smear of organ and orchestral sounds, resulting in undulating timbres that slip effortlessly between following and leading each other. Later, there are much shorter tracks, such as ‘Fundamental’ and ‘Finite Resources’, each under two minutes, that feel like ephemeral études for solo instruments, presenting feasibly alluring soundworlds that close just as they've hooked the listener – a poised blend of the sounds of Klaus Lang, late Morton Feldman and members of the Wandelweiser collective.

So, to the composer's credit, Mapping the Valleys of the Uncanny does indeed sound like contemporary classical music. From what I understand of the press release and accompanying interview, this is what Karlsson set out to do. But to my ears it does so in a sort of removed way, because, well, I am listening to something that is intentionally designed to be only like contemporary classical music. Its construction is so mechanical that it feels bare, not unnerving as other examples of the uncanny might.

It seems, then, that Karlsson is very successful in his endeavours. Perhaps so successful that I feel compelled to consider the reasons behind uncanny music altogether. I can see why, theoretically, it might be interesting to make something explicitly limited to an image of its origin, but to listen to and try and have a relationship with, I personally find it much harder. Maybe the affects of the uncanny don't translate between media: in a lot of visual art and robotics, I find it disconcerting, but here it comes across as automated and – I think this is the right term – soulless. Raising these sorts of questions is no small feat and is a credit to the disc.

But as a listener, as opposed to a musicologist, Mapping the Valleys of the Uncanny makes me miss messiness: the vulnerability of a composer, at some stage and to some degree, having to hold their hands up and say, ‘I did it because I thought it sounded good’. I think it's these decisions that distinguish pieces that I like from those that I love. Karlsson's removal of self is perhaps too effective: I can't hear him.

I wonder if this removal of messiness – of people – within the creative process is most apparent through the overall structure of the album. Mapping the Valleys of the Uncanny clocks in at just under two hours and reasons for this grouping of pieces are hard to decipher. There doesn't seem to be a path plotted that I might be (emotionally) guided down or resist, nor anything resembling either of these as further deployments of the uncanny. But, then again, if committed to the uncanny in method and output, then maybe my expectations of curation aren't appropriate.

The uncanny extends into aspects of the release beyond the sounds. In the accompanying interview, Karlsson claims that through working electronically with algorithms, he's had access to sounds and structures that he wouldn't have had otherwise, due to tradition, tastes and (I think) more practical things like funding. As such, he positions the disc as a sort of political act, (re)claiming new music as his own. While these ideas might have some truth to them, they are common tropes of ongoing conversations concerning music and higher education. His words sound like discussions around accessibility, but they don't call for anything that isn't already in the surrounding rhetoric. Indeed, algorithmic composition is a centuries-old practice which is firmly grounded in (Western) universities. To deploy such techniques suggests a level of (likely institutionalised) education that would parallel what one might need to be interested in and have access to instrumental new music. This suspicion is confirmed by the fact that the music and accompanying book form part of Karlsson's master's degree at the highly selective Royal College of Music in Sweden. If taken not to be uncanny but instead sincere, I worry how his claims might be read by someone who really hasn't had access to new music or anything equivalent for whatever reason. Perhaps a consideration of the ethics of the uncanny is provoked here.

Daniel M. Karlsson's Mapping the Valleys of the Uncanny is a piece of research, then. The disc is good at what it does – that is, creating uncanny new music – which makes me consider the implications of what it's doing altogether. How does the uncanny operate in the realm of new music? Can there be an uncanny composer? What kinds might exist? Possibly important questions, but ones that distract from the experience of listening to the work.